 | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
Should include some reference to W. Stephen Woodward's optoisolator wattmeter which is apparently based on this topology. A brief description is available at http://electronicdesign.com/Articles/Index.cfm?ArticleID=6411. Efadae (talk) 21:05, 15 April 2009 (UTC)
What's a four-quadrant multiplier? --Abdull (talk) 11:14, 22 June 2008 (UTC)
- Ah, Analog multiplier says it all. --Abdull (talk) 11:17, 22 June 2008 (UTC)
This article mentions that a Gilbert cell would not generally be used in an RF mixer. I am studying a textbook by Allen A. Sweet titled "Designing Bipolar Transistor Radio Frequency Integrated Circuits" which suggests that the Gilbert cell is in fact used for RF mixer circuits. Is there a am argument or citation for the suitability of a Gilbert call in RF circuit design? Sevalecan (talk) 05:06, 3 January 2020 (UTC)
Convention vs. history
editThe article is not about the Gilbert cell, but about an earlier (1963) multiplier concept invented by Howard Jones in 1963. They key difference is that the Jones cell multiplies voltages (very small, <100 mV, or far less to keep distortion under control). The Gilbert cell multiplies currents.
 |
 |
|
| Howard Jones, 1963 | Gilbert, 1968 (beta independent, rarely used) | Gilbert, a few years later (beta dependent, more common in use) |
Curiously, many modern textbook attribute the Jones cell to Gilbert, too. Jones' contribution disappears, Gilbert receives no credit for his own contribution. Gilbert himself tried to straighten things up, but eventually gave it up (see page 44 here). Convention and convenience vs. historic truth.
Retired electrician (talk) 14:19, 16 November 2012 (UTC)
- So why not correct the article...
Kevin Aylward 13:40, 7 July 2013 (UTC)
Schematics Disjointed from Text
editThe schematic of the Jones cell is fairly straightforward, but the text implies that Q2/Q6 do not form a differential pair, when in fact they do form one. Perhaps the first sentence of the relevant paragraph intends to say otherwise, but it instead implies that Q1/Q4 and Q3/Q5 are separate stages, while in reality they are separate branches of the same stage. In the process the paragraph fails to mention that the inputs of the Q1/Q4 and Q3/Q5 pairs are connected to the same differential input: for me this is not such a problem, but for the sight-impaired and people who are sufficiently new it can render the workings of the circuit completely incomprehensible.
The two schematics of the Gilbert cell are worse. They appear to show a set of four current sources and six transistors, each schematic with the two outermost BJTs configured as constant-current diodes, but little of this is indicated. Instead, two "diodes" are mentioned, with no indication (including inlined links) that they exist in the form of two of the transistors, where exactly in the schematic they can be found, or how to find them. Further, no clarity is provided as to whether the four current sources are intended to be provided by the cell, or by external circuitry, leaving those who (like me) have SOME understanding of moderately advanced subjects without enough information to be certain that we properly understand what we are looking at.
tl;dr: The "Function" section needs better wording, more explanation, and some inlined links.